AI search is transforming discovery and media economics

Search remains a primary way for publishers to reach audiences. But a growing share of searches now end without a click. Users increasingly find answers directly on results pages or within AI-generated responses. The result is a structural decline in referral traffic and a shift in how discovery works.

This is not only a platform shift. It reflects changing audience behavior, as users turn to AI tools for direct answers. Search is also evolving into an environment where answers are delivered in place.

New research from FT Strategies, The Future of Discovery, examines what this means for publishers. Over time, the industry has moved from owning distribution to relying on platforms to reach audiences. As those platforms change, publishers have less control over discovery and engagement. The question is no longer just how to reach audiences, but how value flows back to content creators.

Content, audience, and value

FT Strategies frames this shift across two dimensions: how content reaches audiences, and what audiences want from it. One axis reflects distribution, from direct relationships to embedded, platform-led access. The other reflects audience needs, spanning information and entertainment. Together, these define four models with different paths forward.

Most publishers operate across more than one model. The challenge is understanding how each contributes to growth and resilience.

1. The niche specialist (direct information) model focuses on high-value content delivered through owned channels. It builds habit through newsletters, apps, and data products. Its advantage lies in depth and trust. Analysis drives engagement beyond what AI summaries provide.

1. The intelligence provider (embedded information) model treats journalism as structured, reusable content. It distributes content into third-party environments, including AI systems. Syndication, APIs, and licensing models tie value to retrieval or citation. This positions publishers as inputs into AI ecosystems.

1. The voice-led brand (direct entertainment) model centers on perspective and identity. Podcasts, newsletters, and personality-driven formats build loyalty. These formats do not rely on algorithms alone. Trusted voices drive repeat engagement.

1. The mass reach publisher (embedded entertainment) model prioritizes scale within platforms. It optimizes visibility across search, social, and aggregators. It offers reach but remains exposed to algorithm changes and traffic volatility.

Risk and reward across models

The four models outline strategic options. The more difficult task is assessing which are viable for a given publisher, and how durable they are as discovery shifts.

FT Strategies introduces assessment models to help quantify these trade-offs. They provide a structured way to test how changes in discovery affect revenue, reach, and long-term resilience. In doing so, they assess how well different models align with a publisher's audience, content, and revenue mix, and how exposed they are to risk and potential return.

• For direct models, the key risk is audience acquisition. As referrals decline, subscription growth depends on fewer channels. Modeling user journeys helps quantify that exposure. It also shows where owned channels can offset it.

• For embedded information models, the risk is substitution. Not all content carries equal value in an AI environment. Content based on proprietary data or expertise is more resilient. Commoditized output is easier to summarize or replicate. This distinction shapes pricing power.

• For platform-driven models, the risk is revenue volatility. Advertising tied to external distribution depends on traffic and visibility. Both can shift quickly. Scenario-based forecasting helps estimate how these changes affect revenue.

Together, these assessments clarify trade-offs between reach, control, and stability. They also highlight where investment is more likely to deliver durable returns.

Discovery becomes part of the audience experience

Discovery now spans the entire audience experience. Content needs to be flexible enough to perform across environments, remain clear when encountered in fragments, and be structured for retrieval. This increases the strategic importance of packaging and distribution.

Control is shifting at the same time. Organizations balance reach through platforms with the stability of owned channels. AI is accelerating this change by reshaping both discovery and value capture. As discovery moves into AI interfaces, value increasingly comes from performance within third-party systems rather than click through.

These dynamics introduce new tradeoffs between reach, attribution, and control. As a result, the economics of publishing are rebalancing. Licensing, structured data, and retrieval-based models are emerging alongside subscriptions and advertising. Many organizations are responding by diversifying across models, aiming to balance reach, ownership, and long-term revenue.